Tritium labeling is an important tool in both clinical and industrial biological laboratories. Its utility has grown rapidly in spite of the technical difficulties encountered. Tritium radiation is extremely difficult to detect and must currently be measured using labor intensive, liquid scintillation techniques which generate large quantities of liquid, radioactive waste. A sensitive, solid state sensor which could measure tritium directly would solve the many of these problems and make possible a very cost effective counting instrument. The Phase I research has clearly demonstrated the feasibility of using a large area avalanche photodiode for this purpose. During Phase II, research will be conducted to further increase the sensitivity of the detector and to assemble a complete working laboratory prototype. A novel 4 element array detector is proposed which will allow the instrument to be both simple to use and cost effective to produce. In this work, we will collaborate closely with the Department of Immunopathology at the Massachusetts General Hospital to develop effective sample preparation protocols and to use the system for a wide variety of samples. By the completion of Phase II, the prototype will have been modified to be optimal for medical use and a successful Phase III commercialization program can be undertaken.